The proposed project takes a population-based approach to neonatal infection and combines information from a variety of care settings - hospital, outpatient office, and emergency room - in order to delineate the population-based epidemiology of infection in the first month of life. Methods will be developed to link automated hospital records of newborns with their HMO's prenatal and post-discharge records to assess the total occurrence of infection during the neonatal period. There are two important goals to this work. First, the algorithm developed will be generally applicable for use in many health care systems to perform efficient, uniformly interpretable, surveillance for newborn infections. Second, this methodology will be used to study important correlates and causes of infection, with the goal of identifying potentially modifiable factors.